Today’s sciences of the artificial extend all the way to the domain of emotions. The recent re-inclusion of the emotional dimension within cognition, in the scientific trend that considers knowing as a form of doing – constructing, creating, fabricating – transforms our understanding of cognitive processes. The science of cognition – which through the development of classical computationalism gave rise to the project of ‘ artificial intelligence ‘ and more recently to that of ‘ artificial life ‘ – has gained a new direction. It is to the project of constructing artificial agents that are able to interact with human actors and to respond to them at an emotional level. This scientific and technological development could transform our world more deeply than have done the internet or cellular telephone which ultimately simply gave us better and different tools and means of communication. For it is not limited to augmenting our capacity of communication, but it promises to extend to new partners our social conversation. Artificial agents are not just instrument, but interlocutors. If these agents prove to be, like human agents, poles of initiative, the consequences of their introduction into our social environment cannot be foreseen. That is why it is fundamental to study the hopes, difficulties and uncertainties, of this development of the science of the artificial while it is still in its (relative) infancy – especially when we take into account the speed at which research progresses.
This domain of research, in both the fields of robotics and of information sciences (where artificial agents present themselves as virtual agents interfacing with humans), is an object of inquiry within a research project centered on the epistemology of emotions. Actually it constitutes an outgrowth of a wider program of research, Empathy at Frontier Sciences, which is financed by the Japanese Society for the Advancement of Science, and whose goal is to explore the possibility of “naturalizing” the concept of empathy in view of recent theoretical and experimental advances, especially in post-classical cognitive science (1).Â The location of this project in Japan, as well as a certain inclination towards constructivist epistemology, suggested dedicating part of the research on the naturalization of empathy to the issue of the artificialization of emotion and of affective relations. At the center of this inquiry is the artificial production of emotion taken as an object research. This raises three fundamental questions. First, is the attempt to generate emotions, affective relations and empathy between human and artificial agents a case of simulation or of creation? This difference is fundamental. It corresponds to the distance that separates making a convincing, and highly sophisticated, puppet from creating an independent agent capable of emotionally interacting with human beings. Second, at the conceptual level, are emotions, empathy, affection purely human characteristics that artifacts imitate or are they something that appears in the interaction itself, and that the use of artificial agent will transform for better or worst? This suggests a third underlying question is: at what point does the simulation or imitation of human behavior cease to be merely imitation to become “the real thing”? Perhaps the difference between the two sides of the alternative presented above is one of degree rather than of nature? These are essentially the issues that we intend to explore through encounters and interview of important researchers in the domain of artificial agents in Japan and other countries. Our method of research is to go visit them in their laboratories, to see and interact with their artifacts, to ask them questions about their research and its intended consequences in their eyes.
The Bassetti Foundation has generously offered us a space on their site to present some of the preliminary results of our research in the form of a research diary. The goal is to present as they happen our encounters with various researchers, their answers to our questions and whenever possible some images of their artifacts, or links to where it is possible to see them. One may wonder, what is the value of presenting such â€˜brute’ results to the public?
In order to have interesting and useful discussions on the ethical import of such research and on the social responsibility that comes with it, it is fundamental first to be aware of what exactly is going on in this field and of the projected social use of such development. At this early stage in our research, this is precisely what we are doing, learning from those who carry out these developments the promises and difficulties involved, the projected use of these artifacts, as well as the goals, and the visions of a future populated with artificial agents, of those who carry them out. It seemed to us that it would be interesting to have a place of ongoing discussion, where we can present the (relatively still brute) result of our encounters with researchers in the field of artificial agents and empathy, in order to receive comments and criticisms and initiate discussions from which we will profit, and where visitors of the Foundation’s website can gain information and participate in shared reflection.
Our first contribution to this blog contains a brief outline of the research project, the protocol of interview that we have been using, and short summaries of our first meeting with researchers.
The project was developed by Luisa Damiano (Ph. D. in Anthropology and Epistemology of Complexity) under the direction of Paul Dumouchel (Professor of philosophy at the Graduate School of Core Ethics and Frontier Sciences of Ritsumeikan University of Kyoto Ritsumeikan University, Kyoto, Japan). (to the text)